2754
SIDNEY IV. Fox AND MILONW. BULLOCK
4.409; mmoles (CaH&B, 2.700, 4.240; in. at. H, 2.513, 3.970; at. H/mole (C4Hs)?B, 0.932,0.936. The concentrated solution of the salt slowly evolved butane, but the rate was too slow for quantitative recovery. At the same time, crystalline material separated. Butane was identified by its msl. wt.: wt. gas g., 0.0302, 0.1010; [CONTRIBUTION FROM
THE
VOl. 73
cc. (S.T.P.), 12.05, 38.57; mol. wt. found, 56.2, 58.7; calcd. for GHIo, mol. wt., 58.1. The weight of the dry Droduct gave no clue as to its comDosition: an analvsb for &hylam&e gave 1.45 moles per moie of original borane. RECEIVEDOCTOBER 22, 1949
PROVIDENCE, R. I.
CHEMICAL LABORATORY O F IOWA
STATE
COLLEGE]
Synthesis of Indoleacetic Acid from Glutamic Acid and a Proposed Mechanism for the Conversion’ BY SIDNEY W. Fox AND MILONW. BULLOCK An economical method for converting glutamic acid to indoleacetic acid, through the intermediate succinaldehydic acid phenylhydrazone, has been developed. a-Ketoglutaric acid has been found as a by-product in the Preparation of the succinaldehydic acid. For the decomposition of the intermediate a-N-chloroaminoglutaric acid there is proposed a mechanism which explains the formation of all the products isolated from the reaction mixture,
Indoleacetic acid has been of interest in a large The synthesis of indoleacetic acid by this innumber of phytological applications which depend expensive conversion of glutamic acid has previously been noted,16but the conditions employed on its function as a growth h ~ r m o n e , heteroauxin. ~-~ Methods of preparation which might increase its a t the time did not give satisfactorily reproducible general availability have been needed. The results.16 This process, illustrated by the reactions various syntheses of indoleacetic acid which may below, has been brought under control ,with details be found in the literature up until 1939 were criti- that are presented in the Experimental section. cally reviewed by Harradence and Lions,6 who NaOCl stressed the inaccessibility of the starting mate- HOOCCHCH~CH~COOH > -. HOOCCHCH~CH~COOH rials. Since that time a number of other syntheses have been The methods XHz I HkCl which have been of most preparative value in- HOOCCHCHzCHzCOOH H f volve the introduction of the acetic acid residue I H NC1 into preformed indole by formaldehyde and hyCHOCHzCHzCOOH HOOCCOCHzCHzCOOH drocvanic acids8through gramine and cvanideSQ by reaction of the indorylGagnesium iodihe comCBH~NHNHZ CeHsNHNHt plex with chloroacetonitrile, and by the reaccsH&“=CHCH,CHzCOOH HOOCCCHzCH2COOH tion of indole with ethyl diazoacetate.11 The I/ principal other route t6 indoleacetic acid has NNHCsH, been through succinaldehydic a ~ i d , ~ ,but ~ l this J~ HzSO4 procedure has suffered from lengthy methods of CzHsOH preparation of the aldehyde. With the availability of glutamic acidlSfrom large-scale manuCHzCOOH facture however, it is possible to prepare succinaldehydic acid14 rapidly, in quantity, and to VLXH/ + OJCOOH convert it into the phenylhydrazone which serves as a convenient method of isolation. The precipiAlthough many of the reactions recorded in the tated oil can then be transformed into indoleacetic above sequence have appeared individually in the acid through the Fischer ring closure. literature, it was necessary in this Laboratory to introduce modifications and integration of steps (1) Journal Paper No. J-1756 of the Iowa Agricultural Experiment Station, Ames, Iowa, Project No. 1110, Chemicals for Agricultural in order to obtain product on a practical scale of Utility. Also Paper V in a series, Amino Acid Conversion Products. operation. Presented before the Division of Organic Chemistry at the American In addition to the succinaldehydic acid formed Chemical Society Meeting, September 8, 1950 a t Chicago. From the in the decomposition of the a-N-chlaroaminoPh.D. thesis of Milon W. Bullock, 1950. (2) F. W. Went and K. V. Thimann, “Phytohormones,” The Macglutaric acid, a small amount of a-ketoglutaric millan Company, Inc., New York, N. Y., 1937. acid also results, as shown above. The resultant F. P. hlehrlich, U. S. Patent 2,428,335; C. A , . 42, 717 (1948). (3) a-ketoglutaric acid phenylhydrazone accompanies (4) G S. Avery, E. B. Johnson, R. hI. Addoms and B. F. Thompson, the desired product and cyclization of the crude “Hormones and Horticulture,” McGraw-Hill Book Co., Inc., New York, N. Y..1947. oil gives some of the 2-carboxyindole-3-aceticacid. (5) R. H. Harradence and F.Lions, Proc. Roy. SOL.N . S . Wales, 7 2 , These two products are easily separated. 221 (1938-1939). Since a-keto acids are not decarboxylated under (6) Z. Tanaka, J . Pharm. SOC.J a p a n , 60, 17 (1940). (7) Z. Tanaka, ibid.,60, 75 (1940). the conditions employed in the decomposition of ( 8 ) K. Bauer and H. Andersag, U. S..Patent 2,222,344. the a-N-chloroaminoglutic acid, the a-keto acid (9) H. R. Snyder and F. J. Pilgrim, THISJOURNAL, 70, 3770 (1948). cannot be an intermediate in the formation of the (10) R. Majima and T . Hoshino, Ber., 68, 2042 (1925). (11) R. W. Jackson and R. H. F. Manske, Can. J . Research, 18B,170 half-aldehyde. To account for the products formed (1935). in the reaction mixture the following mechanism (12) A. Ellinger, Bcr., 87, 1801 (1904). f
4
+
4
+
/\--7~~z~~~~
(13) S. W.Fox, Can. Mcd. Assn. J . , 66, 76 (1947). (14) K.Langheld, Ber., 44, 2371 (1909).
J. Stevens and S. W.Fox, THISJOURNAL, 70, 2283 (1948). (18) F. J. Stevens, P h . D thesis, Iowa State College, 1947. (15) F.
ACID FROM GLUTAMIC ACID INDOLEACETIC
June, 1951
for the decomposition of a-N-chloroaminoacids is proposed
2755
an01 were refluxed five hours under nitrogen. The cooled solution was poured into 11. of water and the oil which sepa-
rated was extracted with three 200-ml. portions of ether. The combined ether extracts were dried over sodium sulfate containing a small amount of potassium carbonate to neutralize any acid in the ether. Distillation of the ether and alcohol left 12.7 g. of an orange oil. The crude ethyl indole-3-acetate was purified by vacuum distillation. The fraction distilling 160-180" at 0.2 mm., wt. 7.2 g., was collected as product. A second fraction distilling 180-220' at 0.2 mm. weighed 1.5 g. Saponification OHof this second fraction gave 0.85 g. of 2-carboxyindole-3acetic acid, which melted 231-233 ' (dec.) after two recrysT e tallizations from water. R-C-COzR-C-CotR-CH COI HCl A n d . Calcd. for CIIHoO,N: neut. equiv., 109.5; N, /I ll 6.40. Found: neut. equiv., 110 (potentiometric): N, HkCl NH C1NH 6.32, 6.34. A mixed melting point with an authentic sample of 2carboxyindole-3-acetic acid, m.p. 236' (dec.) , prepared by, the method of King and L'Ecuyer,l* melted 235-236 R-COCOOH RCHO (dec.). Acidification of the sodium salt of an a-NThe principal product was saponified by refluxing one chloroamino acid (I) gives the free acid which can hour with 100 ml. of 10% methanolic KOH. The solution diluted with 200 ml. of water t n d distilled until the form a six-membered chelated ring (11). If the was The alkaline solutemperature of the vapor reached 99 electron shifts indicated by the arrows occur then tion was extracted once with 50 ml. of ether. Most of the the molecule can decompose to form three small ether was distilled out of the aqueous layer by reducing the stable molecules. Under the conditions of the pressure with an aspirator. Acidification of the aqueous solution with 10% HCI gave a light brown crystalline prodreaction the imine hydrolyzes to form the aldehyde. uct. The crystals were filtered off, washed three times with The yield of the a-keto acid appears roughly to be water and air-dried in the filter. The product was washed proportional to the time elapsed between the with a small amount of chloroform, dried, and recrystallized formation and acidification of the a-N-chloroamino from water (Norit A). This gave 1.82 g. of indoleacetic acid, m.p. 164165" (dec.). A mixed melting point with compound (I). This suggests that a proton is a? sample of indoleacetic acid showed no depresremoved from the salt of the a-N-chloroamino sion.authentic The mother liquors were combined and extracted acid to form the carbanion (111), which immediately with three 30-1111. portions of ether. The combined ether or simultaneously stabilizes itself by the loss of a extracts were dried over sodium sulfate and distilled. The was extracted with 10 ml. of chloroform. The chloride ion. The imino compound may then residue crystals were filtered off and washed with three 5-ml. porbe hydrolyzed. This proposal is in harmony with tions of chloroform. This gave 1.47 g. of product, m.p. the original mechanism suggested by Langheld,14 155-160". After one recrystallization from 20 ml. of water for the formation of aldehydes from a-amino this product melted 165-167" (dec.), and weighed 1.21 g. The total yield of pure indoleacetic acid was 3.03 g. (21%). acids. The results of some other runs are summarized in Table I . 0
4
-
1 +
+
+
.
Experimental
Succinaldehydic Acid Phenylhydrazone .-To 24.9 g . (0.20 mole) of glutamic acid (General Mills) in 400 ml. of 0.5 N sodium hydroxide solution, 200 ml. of fresh 1 M sodium hypochlorite solution17 was added. The alkaline solution was stirred until it gave a negative test with starch-iodide paper, and acidified by the addition of 70 ml. of 3 N HC1 which resulted in evolution of carbon dioxide. The solution was heated to 50" and maintained a t that temperature until the starch-iodide test was negative (50 minutes). A solution of 22.0 g. (0.20 mole) of phenylhydrazine in 50 ml. of 25% acetic acid was added and the reaction mixture was heated for 20 minutes on the steam-bath. The reaction mixture was acidified to Congo red with hydrochloric acid, cooled, and extracted with 200 ml. of ether and with 100 ml. of ether. The combined ether extracts were extracted with 300 ml. of N NaOH. The basic solution was acidified with 10% hydrochloric acid. The phenylhydrazone was extracted with two 100-ml. portions of ether. The combined ether extracts were dried over sodium sulfate and distilled. The yield of crude succinaldehydic acid phenylhydrazone was 18g. (47%). Other yields were in the range of 53-63 %, determined as weighable crystalline p-nitrophenylhydrazone prepared from the aqueous solution. A 67% yield was obtained when Chloramine T was used instead of hypochlorite.16 Indoleacetic Acid.--Sixteen grams (0.083 mole) of crude succinaldehydic acid phenylhydrazone and a solution of 20 ml. of concentrated sulfuric acid in 180 ml. of absolute eth(1%) Prepared by the method of F. Raschig, B e . , 40, 4580 (1907). A commerdal solution of sodium hypochlorite, "Chlorox." was found to be unsatisfactory.
YIELDS OF
Crude succinaldehvdic acid phenylhydrazone, Run
B.
TABLE I ACIDI N
INDOLEACETIC
Cyclization agent, HnSO4, EtOH, ml ml.
.
SEVERAL
RUNS
Yield of DrOduct * Pure indoleCrude, acetic % acid, %
1 28.0 70 700 34 28' 2 9.0 10 90 32 lgb 10 90 32 24b 3 8.2 The ester was distilled in a molecular still. The ester was distilled through a short vacuum-jacketed Claisen still head.
Attempts to cyclize succinaldehydic acid phenylhydrazone with boron trifluoride without solvent, and with acetic acid as solvent, were not successful. An attempt to cyclize the ethyl ester of the succinaldehydic acid phenylhydrazone with boron trifluoride was also unsuccessful, as was an attempt to cyclize the acid with an hydrochloric acid solution saturated with zinc chloride. Since no starting material was isolated in any of the unsuccessful experiments, it is not known whether cyclization failed or whether the negative results were due to the difficulty of isolating the product from the reaction mixture. In none of these unsuccessful experiments, however, was the crude ester from the cyclization mixture purified by distillation before saponification. AMES,IOWA RECEIVEDOCTOBER 26, 1950 (18) F. E. King and P. L'Ecuyer, J . Chcm. Soc., 1901 (1934).
